The role of body mass index, weight change desires

Psychology & Health

ISSN: 0887-0446 (Print) 1476-8321 (Online) Journal homepage: http://www.tandfonline.com/loi/gpsh20

The role of body mass index, weight change desires and depressive symptoms in the healthrelated quality of life of children living in urban disadvantage: Testing mediation models Ciara Wynne, Catherine Comiskey & Sinéad McGilloway To cite this article: Ciara Wynne, Catherine Comiskey & Sinéad McGilloway (2015): The role of body mass index, weight change desires and depressive symptoms in the health-related quality of life of children living in urban disadvantage: Testing mediation models, Psychology & Health, DOI: 10.1080/08870446.2015.1082560 To link to this article: http://dx.doi.org/10.1080/08870446.2015.1082560

Accepted online: 14 Aug 2015.Published online: 11 Sep 2015.

Submit your article to this journal

Article views: 17

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=gpsh20 Download by: [37.228.230.11]

Date: 20 September 2015, At: 08:50

Psychology & Health, 2015 http://dx.doi.org/10.1080/08870446.2015.1082560

The role of body mass index, weight change desires and depressive symptoms in the health-related quality of life of children living in urban disadvantage: Testing mediation models Ciara Wynnea*, Catherine Comiskeya and Sinéad McGillowayb a

School of Nursing & Midwifery, Trinity College Dublin (TCD), Dublin, Ireland; bDepartment of Psychology, National University of Ireland Maynooth (NUIM), Kildare, Ireland

Downloaded by [37.228.230.11] at 08:50 20 September 2015

(Received 7 November 2014; accepted 10 August 2015) Objectives: This study was undertaken to ascertain whether or not the body mass index (BMI) of urban disadvantaged children indirectly affects their health-related quality of life (HRQoL) through weight change desires and depressive symptoms and whether such mediation is conditional upon age and gender. Design: A total of 255 children aged 7–12 years (50% male) were recruited from 7 schools in urban disadvantaged districts in Ireland using consecutive sampling. A prospective longitudinal design was employed whereby children completed, at two time points, the Kidscreen-27, the Children’s Depression Inventory, and the Health Related Behaviour Questionnaire, and had their BMI measured. The analyses involved multiple-, half-longitudinal- and moderated-mediation. Results: Results showed that the depressive symptoms of children wanting to change their weight may have lead, in large part, to poorer HRQoL (specifically psychological well-being when considering longitudinal data) rather than weight status per se. The mediation effect of weight change desires occurred regardless of age or gender. Conclusions: Childhood obesity programmes that traditionally focus on the negatives of obesity and the need to control weight may need to take a more positive approach to health and well-being by, for example promoting intuitive eating, an active lifestyle, body acceptance and good mental health. Keywords: body mass index; body image; children; depressive symptoms; health-related quality of life; mediation models

Introduction The risk of overweight status to the health-related quality of life (HRQoL) of children in disadvantaged populations is attracting increasing concern (Manios & Costarelli, 2011; von Rueden, Gosch, Rajmil, Bisegger, & Ravens-Sieberer, 2006; Wynne et al., 2014). HRQoL refers to a subjective evaluation of multiple quality of life dimensions relating to health (Taylor, Gibson, & Franck, 2008). Whilst many studies have demonstrated small to large effects in the difference between the HRQoL of overweight vs. normal weight children, not all those overweight children have clinically poor HRQoL *Corresponding author. Email: [email protected] © 2015 Taylor & Francis


2

C. Wynne et al.

or significantly poorer HRQoL than their counterparts (Tsiros et al., 2009; Wynne et al., 2014), and there has been some uncertainty regarding the direction of causality (Herman, Hopman, & Craig, 2010; Williams et al., 2011). Consequently, attention has turned to mediators or mechanisms by which overweight status can affect HRQoL, but there is a dearth of research exploring the causal effect of body mass index (BMI) on HRQoL in disadvantaged and other populations. Indirect evidence from the literature suggests that body dissatisfaction (i.e. negative cognitive appraisals and emotions about one’s weight and shape; an aspect of the body image perception construct (Thompson, Heinberg, Altabe, & Tantleff-Dunn, 1999), and depression, which have reached unprecedented levels in young people (O’Connell, 2012; Wesselhoeft, Sørensen, Heiervang, & Bilenberg, 2013), may be key mechanisms (e.g. Chaiton et al., 2009; Kurth & Ellert, 2008), although, this causal effect may be conditional upon age and gender. More specifically, research shows that body dissatisfaction is common in overweight children, though not all of them are aware that they are overweight (Edwards, Pettingell, & Borowsky, 2010; Jansen, van de Looij-Jansen, de Wilde, & Brug, 2008). Body dissatisfaction is found to be associated with lower HRQoL, and it has been suggested that body image perception is also a better indicator of HRQoL than BMI in children (Edwards, Patrick, Skalicky, Huang, & Hobby, 2012; Haraldstad, Christophersen, Eide, Nativg, & Helseth, 2011). Depression is less common in overweight children than body dissatisfaction, such that the association between BMI and depressive symptoms is the modest at best and sometimes negligible (Wardle & Cooke, 2005). However, body dissatisfaction is positively associated with depression, and the HRQoL of children with depression is typically reduced (Bastiaansen, Koot, Ferdinand, & Verhulst, 2004; Paxton, Neumark-Sztainer, Hannan, & Eisenberg, 2006; Stevanovic, 2012). High levels of body dissatisfaction may be due, at least in part, to the internalisation of a ‘body perfect ideal’ from a young age, such as the thin female physique or the lean and muscular male body that are often idealised in Western culture (Grogan, 2007; Schaad, 2012). Thus, if children experience ongoing negative appraisals of their body, and/or perceive these to be uncontrollable, this could lead to depression (Abramson, Seligman, & Teasdale, 1978; Beck, 1967), and especially in young females who tend to mature physically more quickly than males and process selfrelated appearance information more deeply and automatically (Hargreaves & Tiggemann, 2004). There is some evidence to suggest that body dissatisfaction mediates the relationship between obesity and depression and that this mediation occurs for adolescents and girls more so than for children and boys (e.g. Chaiton et al., 2009; Erickson, Robinson, Haydel, & Killen, 2000), but evidence of such conditional mediation is sparse and inconclusive. Moreover, we were unable to locate any moderated mediation studies exploring if body dissatisfaction and subsequent depressive symptoms rather than overweight status, lead to impaired HRQoL in children, and particularly adolescents and girls. The objectives of this study were to determine, in a sample of urban-disadvantaged children: • what the strength and direction of the relationships are between BMI, desires about weight change (an aspect of the body image perception construct), depressive symptoms and HRQoL;

Psychology & Health

3

• whether weight change desires and, in turn, depressive symptoms mediate the relationship between BMI and HRQoL when using cross-sectional and longitudinal data, and; • whether a mediation effect by weight change desires in the relationship between BMI and HRQoL is conditional upon gender and age using cross-sectional data. Method


Sample and procedure The study involved a prospective longitudinal cohort design to assess the height, weight, health and well-being of 255 children aged 7–12 years (50% male) during a typical school day and at two time points 12 months apart; time one and two. At time two, the attrition rate was 22% as 56 children were not available to take part (e.g. they graduated to another school). Children were recruited using consecutive sampling across seven schools in highly disadvantaged urban districts of the Greater Dublin Area in Ireland according to the Government and the Deprivation Index (Department of Education & Science, 2003; Haase & Pratschke, 2008). This study was part of a project which was evaluating the initial implementation of a Health Promoting Schools model in these schools (Comiskey et al., 2012). Ethical approval was obtained from the Faculty of Health Sciences Ethics Committee in Trinity College Dublin. Parents/guardians provided informed consent for their children whilst assent was also obtained from children. Measures Adiposity was determined by BMI [weight/(height)2] which was measured by children’s nurses trained in anthropometric measurement (Hollywood et al., 2012). AnthroPlus software was used to convert BMIs into age- and gender-specific standardised Z-scores ranging between −3 and 3 and to identify weight categories using WHO (2007) thresholds for overweight (i.e. +1 SD or 25 kg/m2) and obesity (i.e. +2 SD or 30 kg/m2) (in de Onis et al., 2007). cChildren completed the Health Related Behaviour Questionnaire (HRBQ); a 31-item, self-report questionnaire for 7–12-year-olds that provides descriptions of a range of health beliefs, behaviours and feelings (Balding, 2008). Only one question in the HRBQ pertained to weight change desires which asks: Which sentence currently describes you best: ‘I would like to put on weight’; ‘I would like to lose weight’; or ‘I am happy with my weight as it is’. For path analysis, the weight change desires variable was categorised into ‘happy with weight’ and ‘wanting to change weight’ (put on/lose combined). The Children’s Depression Inventory Short-form is a ten-item self-report clinical objective measure of depressive symptoms in children aged 7–17 years (Kovacs, 2009). For every item, respondents choose one of the three statements that best describes their feelings over the past two weeks. T-scores with a mean of 50 (SD = 10) are produced and scores >65 are indicative of clinically significant depression. The instrument demonstrates strong psychometric properties and good reliability with a Cronbach’s α of 0.80 (Kovacs, 2009; Meehan, Houghton, Cowley, Houghton, & Kelleher, 2008). In the current study α was .82. HRQoL was assessed using the self-report Kidscreen-27. This is a 27-item generic (i.e. as opposed to disease specific) measure of HRQoL for children aged 8–18 years

4

C. Wynne et al.


(Kidscreen Group Europe, 2006). It assesses five dimensions of well-being using a five-point Likert scale (e.g. 1 = never and 5 = always). These are: (1) physical wellbeing; (2) psychological well-being; (3) autonomy and parent relations; (4) social support and peer relations and (5) school environment. A total HRQoL score can be calculated from 10 of the items which correspond to those in the shorter Kidscreen-10 measure. T-scores range from 0 to 100, with a mean of 50 (SD = 10). Higher scores indicate better HRQoL. The Kidscreen-27 and Kidscreen-10 are psychometrically robust with Cronbach’s α over .73 in the literature (Kidscreen Group Europe, 2006; RavensSieberer et al., 2007). In the present study, α for total HRQoL was satisfactory at .81. Statistical analysis Pearson’s r and point-biserial correlations were used to test the bivariate association between all variables. Multiple mediation was conducted to test whether weight change desires and, in turn, depressive symptoms total T-scores mediate the relationship between BMI Z-scores and HRQoL total T-scores at time one (model 1a) (Hayes, Preacher, & Myers, 2011). Model 1a was explored further using half-longitudinal mediation (Cole & Maxwell, 2003) to determine whether weight change desires mediate the relationship between BMI Z-scores and HRQoL total T-scores between time one and two (model 1b), and whether depressive symptoms total T-scores mediate the relationship between weight change desires and HRQoL total T-scores between time one and two (model 1c). These two models were employed to investigate longitudinal mediation across multiple mediation paths as half-longitudinal mediation cannot be conducted on more than one mediator at a time. In half-longitudinal mediation analysis, data at time two were assumed to remain constant at a hypothetical third time point: time three. Moderated mediation (Preacher, Rucker, & Hayes, 2007) was conducted to determine whether a mediation effect by weight change desires in the relationship between BMI Z-scores and HRQoL total T-scores is conditional upon gender at time one (moderated between the path from BMI to weight change desires) (model 1d), and conditional upon age (grouped into 7–10 and 11–12 years) at time one (moderated between the path from BMI to weight change desires) (model 1e). The non-parametric bootstrapping method was used to calculate the standardised betas (β) of the direct-, mediation/indirect- and total- effects and the bias corrected (BC) confidence intervals (CIs, 95%) of models 1a–1e (Preacher & Hayes, 2004). A bootstrapped sample of N = 5000 was used. While some distributions of the study outcome were non-normal and skewed, this method does not require normally distributed data (MacKinnon, Lockwood, & Williams, 2004). In the analysis, missing data were excluded pairwise. Thus, the tables report actual response rates and valid percentages. All analyses were conducted with Mplus version 7 Muthén and Muthén (1998–2012). According to simulated power tables, a sample size of 200–500 would be needed to detect moderated mediation with 80% power using bootstrapping (Chu & Chen, 2012). Results Sample descriptives At time one, over one third (38%, 96/253) of children were considered either overweight (16%) or obese (22%) (Table 1), which exceeds the 25 and 33% rates observed


Psychology & Health

5

in nationally representative (Whelton et al., 2007) and socioeconomically disadvantaged (Layte & McCrory, 2011) samples of Irish children, respectively. With regard to weight change desires, nearly two thirds of children were happy with their weight (61%), 5% would have liked to gain weight and approximately one third would like to lose weight (34%). The last of these is similar to the proportion of 10–17-year-olds from a nationally representative sample in Ireland (32%) who reported wanting to lose weight (Kelly, Gavin, Molcho, & Nic Gabhainn, 2012). Furthermore, over two fifths (43%) of children assessed as overweight and obese, were happy with their weight, while 26% of normal weight children wanted to change their weight. Whilst the mean total T-scores for depressive symptoms fell within the average range by international standards (Kovacs, 2009), the proportion of children who reached clinically significant levels of depression was 7%; this is lower than the proportion (13.6%) identified in the general Irish population of 6–18-year-old children (Houghton, O’Connell, & O’Flaherty, 1998). Finally, the mean HRQoL total T-scores was within the average range for the general population of Irish and European children (Kidscreen Group Europe, 2006). Bivariate analysis As expected, those who wanted to change their weight had significantly higher levels of body fat and depressive symptoms, and significantly lower levels of total HRQoL than those who were happy with their weight, with small and medium effect sizes Table 1. Sample descriptives at year one. Variables Age Gender Male Female BMI Z-scores Underweight a Normal weight Overweight Obese Weight change desires Happy with weight Wanting to lose weight Wanting to gain weight b Depressive symptoms total T-scores Non-cases Cases HRQoL total T-scores Physical Well-being T-scores Psychological Well-being T-scores Autonomy and Parent Relations T-scores Social support and Peer Relations T-scores School Environment T-scores a

N (%)

M (SD) range

255 (100) 255 (100) 128 (50) 127 (50) 253 (99) 1 (.4) 156 (61.6) 41 (16) 55 (22) 252 (99) 154 (61) 85 (34) 13 (5) 252 (99) 235 (93) 17 (7) 245 (96) 253 (99) 252 (99) 252 (99) 251 (99) 251 (99)

9.73 (1.46) 7–12

0.84 (1.10) −2.21–4.41

47.17 (9.86) 40–94 53.86 53.57 53.56 52.07 53.26 55.47

(12.03) (11.05) (10.98) (12.88) (13.25) (11.86)

26.64–83.81 20.70–73.20 20.55–75.53 13.96–74.39 11.24–66.34 16.28–71.00

As only 1% of children were underweight they were not analysed separately in further analysis. As only a small proportion of children reported wanting to gain weight for all further analyses weight change desires is a dichotomous variable; happy with weight and wanting to change weight. b

6

C. Wynne et al.


observed (Table 2). BMI Z-scores were also significantly negatively correlated with HRQoL total T-scores, showing a small effect size. In addition, depressive symptoms total T-scores were significantly negatively correlated with HRQoL total T-scores, with a large effect size. No correlation was found between depressive symptoms total T-scores and BMI Z-scores. Gender and age were not found to confound any of the above relationships. Arguably, all study variables within the path analysis should correlate significantly with each other as a prerequisite to mediation analysis (Baron & Kenny, 1986; Judd & Kenny, 1981). However, in line with Preacher and colleagues (2007), this study holds that an indirect effect can exist even if some study variables are not significantly correlated. Mediation models Significant mediation exists when the indirect path coefficient is significantly different from zero or when zero is not contained within the upper and lower BC CIs. Model 1a: multiple mediation Table 3 shows that zero was not contained within the BC limits of the significant indirect effect (a1b3b2) found between BMI Z-scores and HRQoL total T-scores. The effect of BMI Z-scores on HRQoL total T-scores (path c) became non-significant when controlling for weight change desires and depressive symptoms total T-scores (path c1). Thus, depressive symptoms total T-scores fully mediated the relationship between BMI Z-scores and HRQoL total T-scores through weight change desires, explaining 51% of the effect (proportion calculated from formula: a1a3b2/(c1+a1b1 + a2b2 + a1a3b2)). In other words, the high depressive symptoms of those who wanted to change their weight may have led to poorer HRQoL rather than a high BMI, whilst the positive mental health of those happy with their weight may explain a positive HRQoL rather than a low BMI (Figure 1). Notably, depressive symptoms total T-scores alone did not mediate the relationship between BMI Z-scores and HRQoL total T-scores. Models 1b and 1c: half-longitudinal mediation Model 1a was explored further using longitudinal data. Table 4 shows that in model 1b where weight change desires is assumed to be the mediating variable, zero was contained within the BC limits of the indirect effect (a2b2) between BMI Z-scores and HRQoL total T-scores (Figure 2). Similarly, model 1c results (Table 4) revealed that zero was also contained within the BC CIs of the indirect effect (a2b2) between weight change desires and HRQoL total T-scores, where depressive symptoms total t-scores is assumed to be the mediating variable (Figure 3). These results indicate that weight change desires did not mediate the relationship between BMI Z-scores and HRQoL total T-scores, and depressive symptoms total T-scores did not mediate the relationship between weight change desires and HRQoL total T-scores when tested over time. However, further analysis of the five dimensions of HRQoL revealed that a mediation effect occurred for both models when psychological well-being T-scores was the dependent variable (Indirect effect a2b2 of model 1b: β = −.878, BC CI = {−2.019, −.210}; Indirect effect a2b2 of model 1b: β = .361, BC CI = {.002, 1.174}).

Age Gender a BMI Z-scores Weight change desires a Depressive symptoms T-scores HRQoL physical well-being T-scores HRQoL psychological well-being T-scores HRQoL autonomy & parent relations T-scores HRQoL social support & peer relations T-scores HRQoL school environment T-scores HRQoL total T-scores

1 .12* −.05 −.07 −.04 −.12 −.05 .13* .07 −.17** .00

2

.09 −.08 −.04 .01 .03 −.09 −.16** −.16* −.03

3

−.34** .07 −.17** −.12 −.13* −.12 −.08 −.15*

4

−.29** .37** .29** .26** .15* .25** .31**

5

−.35** −.63** −.44** −.54** −.45** −.55**

6

.52** .38** .42** .47** .68**

7

.55** .55** .52** .75**

8

.53** .42** .72**

9

.48** .65**

10

.67**

Notes: aPoint-biserial correlation was used for dichotomous variables. Effect size is small when r = ± .10 to ± .29; medium when r = ± .30 to ± .49; and large when r = ± .50 to ± 1.0. *p < .05; **p < .01.

1 2 3 4 5 6 7 8 9 10 11

Variables

Table 2. Pearson’s r correlations between study variables.


Psychology & Health 7

8

C. Wynne et al.

Table 3. Path coefficients and indirect effect confidence intervals for model 1a. Model 1a

p

−.413 −.106 −.413

.000** .173 .000**

.223 −.460

.021* .000**

−.152 −.031 −.078

.014* .620 .001**

Notes: Testing model 1a involved the following regression equations: M1T1 = a1XT1 + eM1T1; M2T1 = a2X1 2 1 2 T1 + a3M T1 + eM T1; and YT1 = b1M T1 + b2M T1 + c1XT1 + eYT1. Model fit statistics were satisfactory (i.e. comparative fit index (CFI) was satisfactorily greater than .90, root mean square error of approximation (RMSEA) and weighted root mean square residual (WRMSR) were good at < .05, while X2 was unexpectedly significant). BC 95% CI for the unstandardised beta coefficient of the indirect effect does not contain zero, which indicates a significant indirect effect. β = standardised estimates, *p < .05; **p < .01.

Weight change desires

BMI z-scores

-.031 (-.152*) c1

HRQoL total t-scores

-.413** a3


BMI Z-scores to Weight change desires (path a1) BMI Z-scores to Depressive symptoms total T-scores (path a2) Weight change desires to Depressive symptoms total T-scores (path a3) Weight change desires to HRQoL total T-scores (path b1) Depressive symptoms total T-scores to HRQoL total T-scores (path b2) BMI Z-scores to HRQoL total T-scores (path c) BMI Z-scores to HRQoL total T-scores (path c1) Indirect effect (path a1a3b2) BC CI {−.118, −.039}**

β

Depressive symptoms total t-scores Figure 1 Standardised coefficients for the multiple mediator effects of weight change desires and depressive symptoms total t-scores in the relationship between BMI z-scores and HRQoL total t-scores at year one (*p < .05, **p < .01).

Model 1d and 1e: moderated mediation Unexpectedly, initial regression analysis for model 1d indicated that the interaction of BMI Z-scores and gender (path a3) did not significantly predict weight change desires over and above the main effects of BMI Z-scores and gender, independently (Table 5). Thus, the indirect effect (a1 + a3 W)(b1) for both genders was not explored any further. Hence, weight change desires may serve as a mediator in the relationship between BMI Z-scores and HRQoL total T-scores, but the strength of the effect may not differ by gender (Figure 4).

Psychology & Health

9

Table 4. Path coefficients and indirect effect confidence intervals for models 1b and 1c.

Model 1b BMI Z-scoresT1 to Weight change desiresT2 (path a2) Weight change desiresT1 to HRQoL total T-scoresT2 (path b2) Indirect effect (path a2b2) BC CI {−1.603, .013} Model 1c Weight change desiresT1 to Depressive symptoms total T-scoresT2 (path a2) Depressive symptoms total T-scoresT1 to HRQoL total T-scoresT2 (path b2) Indirect effect (path a2b2) BC CI {−.054, .700}


M1T

β

p

−.267 .089 −.593

.007** .157 .220

−.053

.359

−.144

.036*

.164

.455

mM1T-1+a1XT−1

Notes: Testing models 1b and 1c involved the following regression equations: = and YT = yYT−1+b1M1T−1+eYT. Model fit statistics were satisfactory for all models. BC CIs of the unstandardised beta coefficients contain zero indicating non-significant effects. β = standardised estimate, *p < .05; **p < .01.

+ eM1T

BMI z-scores T1

Weight change desires T1

HRQoL total t-scores T1


y



Figure 2 Standardised coefficients for the relationship between BMI z-scores and HRQoL total t-scores as mediated by weight change desires between year one (T1) and two (T2) (Assuming stationarity, the half-longitudinal mediation design predicts that path b2 equals path b3, and that the indirect effect a2b2 is equivalent to a2b3. T3 refers to a hypothetical time three (**p < .01)).

Similarly, in model 1e the interaction of BMI Z-scores and age (path a3) did not significantly predict weight change desires over and above the main effects of BMI Z-scores and age, independently (Table 5). Thus, weight change desires may serve as a mediator in the relationship between BMI Z-scores and HRQoL total T-scores regardless of age (Figure 5). Discussion This study investigated the extent to which – in a sample of urban-disadvantaged children – weight change desires and, in turn, depressive symptoms mediated the relation-

10

C. Wynne et al.



Depressive Symptoms total t-scores T1

Depressive Symptoms total t-scores T2




Figure 3 Standardised coefficients for the relationship between weight change desires and HRQoL total t-scores as mediated by depressive symptoms total t-scores between year one (T1) and two (T2) (*p < .05). Table 5. Path coefficients and indirect effect confidence intervals for models 1d and 1e.

Model 1d BMI Z-scores to Weight change desires (path a1) BMI Z-scores × Sex to Weight change desires (path a3) Weight change desires to HRQoL total T-scores (path b1) BMI Z-scores to HRQoL total T-scores (path c) BMI Z-scores to HRQoL total T-scores (path c1) Indirect effect for Sex (female) (a1 + a3 W)(b1) BC CI {−2.631, −.660}* Indirect effect for Sex (male) (a1 + a3 W)(b1) BC CI {−3.459, −1.066}* Model 1e BMI Z-scores to Weight change desires (path a1) BMI Z-scores × Age to Weight change desires (path a3) Weight change desires to HRQoL total T-scores (path b1) BMI Z-scores to HRQoL total T-scores (path c) BMI Z-scores to HRQoL total T-scores (path c1) Indirect effect for Age (older) (a1 + a3 W)(b1) BC CI {−5.555, −1.630}* Indirect effect for Age (younger) (a1 + a3 W)(b1) BC CI {−2.231, −.574} *

β

P

−.603 .207 .417 −.152 .128 −1.474 −2.048

.025* .434^ .000** .014* .543 .015* .015*

.139 −.602 .415 −.152 .067 −3.132 −1.267

.632 .071^ .000** .014* .724 .011* .011*

Notes: Testing model 1d and 1e involved the following regression equations: M1T1 = a1XT1 + a2W1T1 + a3M1T1W1T1 + eM1T1 and YT1 = b1M1T1 + c1XT1 + c2W1T1 + c3XT1W1T1 + eYT1. Model fit statistics were satisfactory for all models. BC CIs for the unstandardised beta coefficients of the indirect effects did not contain zero, which indicates significant indirect effects. ^However, sex and age did not significantly moderate the relationship between BMI Z-scores and Weight change desires. Hence, conditional mediation did not occur. β = standardised estimates, *p < .05; **p < .01.

ship between BMI and HRQoL, and if this was conditional upon age and gender. The findings indicate that high depressive symptoms in those children who wanted to change their weight, may have led to poorer HRQoL rather than a high BMI, whilst the positive mental health of those who were happy with their weight, may explain a

Psychology & Health

11

.207 a3


BMI z-scores

HRQoL total t-scores

Sex

c3

BMI zscoresxSex

Figure 4 Standardised coefficients for weight change desires as a mediator in the relationship between BMI z-scores and HRQoL total t-scores moderated by gender at year one (*p < .05, **p < .01).


-.602 a3


c2

BMI z-scores

HRQoL total t-scores c2

Age

c3

BMI zscoresxAge

Figure 5 Standardised coefficients for weight change desires as a mediator in the relationship between BMI z-scores and HRQoL total t-scores moderated by age at year one (**p < .01).


12

C. Wynne et al.

positive HRQoL rather than BMI in the normal range. This was the case for total HRQoL and all dimensions when cross-sectional data were explored and for psychological well-being when longitudinal data were analysed. The findings also showed that the mediation effect was not moderated by age or gender. Although this is, to the best of our knowledge, the first study to examine these mediation paths, indirect evidence from the literature supports the findings (e.g. Chaiton et al., 2009; Edwards et al., 2010; Erickson et al., 2000; Kurth & Ellert, 2008; Rajmil et al., 2009; Wardle & Cooke, 2005). For example, the finding here that not all overweight children recognise their weight status (i.e. over two fifths of overweight children were happy with their weight) is supported by previous work (Brener, Eaton, Lowry, & McManus, 2004; Kurth & Ellert, 2008) including a nationally representative study of American adolescents (n = 12,853) which showed that 29–33% of overweight adolescents perceived their weight to be higher or lower than their actual weight (Edwards et al., 2010). Moreover, Wardle and Cooke’s (2005) review of 18 studies (1995–2004) indicated that overweight children and adolescents are not necessarily depressed, which also reflects the findings reported here. Further evidence from a large German study of 11–17-year-olds (N = 6669) supports the finding that weight change desires mediate the relationship between BMI and HRQoL; that is, those who were normal weight but concerned about their body size, reported significantly worse HRQoL than their overweight counterparts who perceived their body size to be just right (Kurth & Ellert, 2008). Additionally, a mediation study by Chaiton and colleagues (2009) which assessed 2294 adolescents (13–16 years) found that adiposity indirectly affected depressive symptoms through body dissatisfaction; this further supports the multiple mediation found in the current study. By contrast, however, these authors found that mediation was conditional upon gender (i.e. girls only). Gender and age differences most probably did not arise in our study as we focused on pre-adolescents and therefore, gender-based pubertal factors were less likely to have an effect. Finally, the findings from our longitudinal analysis are also similar to those reported by Rajmil et al. (2009) who found that deteriorating mental health in a general population sample (N = 454; 8–18 years) predicted the most severe deterioration in half of the HRQoL dimensions and particularly psychological well-being. It is possible that some of the children in this study internalised a thin ideal from a young age allowing it to form part of their value system; thus, these children and especially those who were overweight, likely found a discrepancy between their actual and ideal body image thereby leading to cognitive dissonance, which may have led, in turn, to a desire on their part to lose weight (Davison, Markey, & Birch, 2003; Festinger, 1954; Festinger, 1957; Musher-Eizenman, Holub, Edwards-Leeper, Persson, & Goldstein, 2003; Rokeach, 1973). Some of these children may have experienced ongoing negative emotions about their weight, perceived it to be uncontrollable, and hence reported more depressive symptoms (Abramson et al., 1978; Beck, 1967). Subsequently, these children may have perceived a gap between their healthy past and present illbeing and between their present and future aspirations for HRQoL (de Leval, 1995), and evaluated their HRQoL as poorer than those who were happy with their weight and who had positive mental health. Alternatively, children of all sizes who reported being happy with their weight, may not have valued a perfect body or recognised their weight status, or may have employed coping strategies to reduce the psychological discomfort

Psychology & Health

13


associated with wishing to change their weight, thereby experiencing a positive HRQoL (Braun & Wicklund, 1989; Cash, Santos, & Williams, 2005; Heinberg, 2005). Strengths and limitations The findings reported here represent an important addition to the literature as they address a gap in knowledge and theory relating to the association between children’s adiposity and HRQoL by extending the research into mediators and conditional indirect effects of this relationship. The use of self-report measures with good psychometric properties and the assessment of children’s BMI by trained nurses helped to ensure that results were reliable and valid. Furthermore, the non-parametric bootstrapping methods employed here are, arguably, superior to other methods including parametric techniques (Cheung & Lau, 2008; MacKinnon et al., 2004). Finally, the incorporation of a longitudinal element strengthened the research design and allowed for an exploration of causal relationships (Page, Cole, & Timmreck, 1995). However, several limitations of the study should be considered when interpreting the findings. Firstly, weight change desires are only one aspect of body image perception and these were assessed using a single-item question with categorical responses from the HRBQ. Thus, reliability and validity may be compromised (Balding, 2008) and less may have been learned than from analysing continuous data, or data obtained using a full measure of either weight change desires or of body image perception (Shroff, Calogero, & Thompson, 2009; Turner, Dobson, & Pocock, 2010). For example, the HRBQ item did not capture whether a desire to lose or gain weight was for reasons related to thinness, muscularity or fitness; it may be possible that boys unsatisfied with their muscle tone could endorse wanting to lose or gain weight. Although it would have been preferable to administer a more comprehensive and psychometrically robust measure of body image perception (e.g. Body Image Assessment or Body Satisfaction Scale, Siegel, Yancey, Aneshensel, & Schuler, 1999; Veron‐Guidry & Williamson, 1996), this was a convenience sample which was part of an existing project which required other outcomes to be measured, and there was a need, therefore, to avoid overburdening the children. Secondly, while half-longitudinal mediation models are more robust than simple mediation models, the assumption of stationarity and that a certain time interval must elapse for one variable to have an effect on another, could not be tested in this study owing to the lack of a third wave of data (Cole & Maxwell, 2003). Violation of these assumptions may have produced misleading indirect effect estimates (Gollob & Reichardt, 1985). Hence, caution is advised when making causal inferences from the findings. Finally, non-randomised sampling was used to recruit children specifically living in highly disadvantaged urban districts of Dublin. This may have increased the risk of selection bias (Berra et al., 2007). An incorrect characterisation of the population could be drawn if results were generalised to non-urban disadvantaged populations, and therefore, caution is advised in so doing. Furthermore, the finding that there was a smaller proportion of children with depression and below average HRQoL, and a similar percentage wanting to lose weight when compared to the general population, was unexpected given that disadvantaged children are twice as likely to become depressed (Melchior, Moffitt, Milne, Poulton, & Caspi, 2007), tend to have poorer HRQoL

14

C. Wynne et al.


(Kidscreen Group Europe, 2006), and are less likely to have negative body image perceptions (Walters & Kendler, 1995) than those who are more affluent. This suggests that these children may not be necessarily typical of those living in areas of urban disadvantage. However, this last finding may signal that, just like obesity, there is an increase in children wanting to lose weight in disadvantaged urban areas. Indeed, research shows that body dissatisfaction is experienced in children as young as four years (Davison et al., 2003; Musher-Eizenman et al., 2003) and has reached unprecedented levels in young people (O’Connell, 2012), as has obesity (Knai, Lobstein, Darmon, Rutter, & McKee, 2012).

Implications The findings have important implications for policy and practice. Research suggests that young people who engage in weight control behaviours do so in an effort to become healthier and happier (Blaine, Rodman, & Newman, 2007). However, the findings of this study and others, indicate that contemplating engaging in weight control behaviours may lead instead to poorer mental health and well-being. Thus, we may need to take a more positive approach to improving children’s health and well-being such as the health at every size approach, which promotes the benefits of healthy eating and exercise, and supports good mental health, rather than just focusing on the negatives of obesity and the need to control weight (Robison, 2005). Children could be encouraged to eat intuitively; that is, eating in a flexible manner for pleasure whilst all the time attending to internal cues of hunger, satiety and appetite (Healy et al., 2015). Moreover, children could be better supported to find joy in moving their body and becoming more physically vital (i.e. active embodiment, Bacon, 2010). There should also be an awareness and recognition that some children are not happy with their weight and that this can lead to depressive symptoms, thereby suggesting a need for intervention. For example, the promotion of positive rational body acceptance strategies (see Cash et al., 2005; Goss & Allan, 2010) has been shown to produce sustained improvements in body image perception and mood, and has even led to weight loss in obese adults even though this was not the focus (Bacon & Aphramor, 2011; Hutchinson & Calland, 2011; Stice & Shaw, 2004). Several reviews of effective ways to reduce depressive symptoms/episodes have also been published (e.g. David-Ferdon & Kaslow, 2008; Horowitz & Garber, 2006), whilst some methods have also been shown to improve HRQoL (e.g. Vitiello et al., 2006). This study could be replicated in the future using a random sample of children to increase the generalisability of the findings to the urban-disadvantaged or the general population. Body image perception could also be assessed more broadly (in underweight and overweight children) whilst a design with at least three waves of data could be used to determine causality. Future research should also investigate further the mechanisms by which adiposity affects children’s HRQoL. In summary, this study found that – in a sample of urban-disadvantaged children – the depressive symptoms of those who wanted to change their weight may have lead, in large part, to poorer HRQoL rather than weight status per se. Likewise, children who are happy with their weight and have a positive mental health, regardless of their size, may have a better HRQoL. These findings imply that we may need to take a more

Psychology & Health

15

positive approach to improving children’s health and well-being, which concurrently promotes intuitive eating, an active lifestyle, body acceptance and good mental health, rather than just focusing on the negatives of obesity and the need to control weight. Acknowledgements The authors extremely grateful to all the parents, guardians, children and school staff for taking part in, and facilitating, this study.

Disclosure statement


No potential conflict of interest was reported by the authors.

Funding This study was supported by the Irish Research Council under the Postgraduate Scholarship Scheme; and the Childhood Development Initiative (CDI). CDI is one of four area-based initiatives in Ireland which are co-funded by the Atlantic Philanthropies and the Department of Children and Youth Affairs.

References Abramson, L. Y., Seligman, M. E., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49. Bacon, L. (2010). Health at every size: The surprising truth about your weight. Dallas: BenBella Books. Bacon, L., & Aphramor, L. (2011). Weight science: Evaluating the evidence for a paradigm shift. Nutrition Journal, 10, 2–13. Balding, J. (2008). Collecting good data: The primary health related behaviour questionnaire (Version 11.2 Manual). Exeter: Schools Health Education Unit. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Bastiaansen, D., Koot, H. M., Ferdinand, R. F., & Verhulst, F. C. (2004). Quality of life in children with psychiatric disorders: Self-, parent, and clinician report. Journal of the American Academy of Child & Adolescent Psychiatry, 43, 221–230. Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects (Vol. 32). Philadelphia: University of Pennsylvania Press. Berra, S., Ravens-Sieberer, U., Erhart, M., Tebé, C., Bisegger, C., Duer, W., … the European, Kidscreen group. (2007). Methods and representativeness of a European survey in children and adolescents: The KIDSCREEN study. BMC Public Health, 7, 182. Retrieved from http://www.biomedcentral.com/1471-2458/7/182 Blaine, B. E., Rodman, J., & Newman, J. M. (2007). Weight loss treatment and psychological well-being: A review and meta-analysis. Journal of Health Psychology, 12, 66–82. Braun, O. L., & Wicklund, R. A. (1989). Psychological antecedents of conspicuous consumption. Journal of Economic Psychology, 10, 161–187. Brener, N. D., Eaton, D. K., Lowry, R., & McManus, T. (2004). The association between weight perception and BMI among high school students. Obesity Research, 12, 1866–1874.


16

C. Wynne et al.

Cash, T. F., Santos, M. T., & Williams, E. F. (2005). Coping with body-image threats and challenges: Validation of the Body Image Coping Strategies Inventory. Journal of Psychosomatic Research, 58, 190–199. Chaiton, M., Sabiston, C., O’Loughlin, J., McGrath, J. J., Maximova, K., & Lambert, M. (2009). A structural equation model relating adiposity, psychosocial indicators of body image and depressive symptoms among adolescents. International Journal of Obesity, 33, 588–596. Cheung, G. W., & Lau, R. S. (2008). Testing mediation and suppression effects of latent variables: Bootstrapping with structural equation models. Organizational Research Methods, 11, 296–325. Chu, A. T., & Chen, Y. C. (2012). The sample size of a moderator mediator effect. Kaohsiung: Institute of Statistics National University of Kaohsiung. Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling. Journal of Abnormal Psychology, 112, 558–577. Comiskey, C. M., O’Sullivan, K., Quirke, M. B., Wynne, C., Kelly, P., & McGilloway, S. (2012). Evaluation of the effectiveness of the childhood development initiative’s healthy schools programme: Dublin: Childhood Development Initiative (CDI). David-Ferdon, C., & Kaslow, N. J. (2008). Evidence-based psychosocial treatments for child and adolescent depression. Journal of Clinical Child & Adolescent Psychology, 37, 62–104. Davison, K. K., Markey, C. N., & Birch, L. L. (2003). A longitudinal examination of patterns in girls’ weight concerns and body dissatisfaction from ages 5 to 9 years. International Journal of Eating Disorders, 33, 320–332. de Leval, N. (1995). Scales of depression, ill-being and the quality of life – Is there any difference? An assay in taxonomy. Quality of Life Research, 4, 259–269. de Onis, M., Onyango, A. W., Borghi, E., Siyam, A., Nishida, C., & Siekmann, J. (2007). Development of a WHO growth reference for school-aged children and adolescents. Bulletin of the World Health Organization, 85, 660–667. Department of Education and Science. (2003). Delivering equality of opportunity in schools. Dublin: Government Publications. Edwards, N. M., Pettingell, S., & Borowsky, I. W. (2010). Where perception meets reality: Selfperception of weight in overweight adolescents. Pediatrics, 125, e452–e458. doi:10.1542/ peds.2009-0185 Edwards, T., Patrick, D., Skalicky, A., Huang, Y., & Hobby, A. (2012). Perceived body shape, standardized body-mass index, and weight-specific quality of life of African-American, Caucasian, and Mexican-American adolescents. Quality of Life Research, 21, 1101–1107. doi:10.1007/s11136-011-0019-0 Erickson, S. J., Robinson, T. N., Haydel, K. F., & Killen, J. D. (2000). Are overweight children unhappy? Archives of Pediatrics & Adolescent Medicine, 154, 931–935. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. Festinger, L. (1957). A cognitive theory of dissonance. Evanston, IL: Row Petersen. Gollob, H. F., & Reichardt, C. S. (1985). Building time lags into causal models of cross-sectional data. Paper presented at the Proceedings of the Social Statistics Section of the American Statistical Association. Washington, DC: American Statistical Association. Goss, K., & Allan, S. (2010). Compassion focused therapy for eating disorders. International Journal of Cognitive Therapy, 3, 141–158. Grogan, S. (2007). Body image: Understanding body dissatisfaction in men, women and children. East Sussex: Routledge. Haase, T., & Pratschke, J. (2008). New measures of deprivation for the Republic of Ireland. Dublin: Pobal.


Psychology & Health

17

Haraldstad, K., Christophersen, K. A., Eide, H., Nativg, G. K., & Helseth, S. (2011). Predictors of health-related quality of life in a sample of children and adolescents: A school survey. Journal of Clinical Nursing, 20, 3048–3056. Hargreaves, D. A., & Tiggemann, M. (2004). Idealized media images and adolescent body image: “Comparing” boys and girls. Body Image, 1, 351–361. Hayes, A. F., Preacher, K. J., & Myers, T. A. (2011). Mediation and the estimation of indirect effects in political communication research. In E. P. Bucy & R. L. Holbert (Eds.), Sourcebook for political communication research: Methods, measures, and analytical techniques (pp. 434–465). New York: Routledge. Healy, N., Joram, E., Matvienko, O., Woolf, S., Knesting, K., Weare, K., & Weare, K. (2015). Impact of an intuitive eating education program on high school students’ eating attitudes. Health Education, 115, 214–228. Heinberg, L. (2005). The role of body image distress in obesity: Is it motivating or malevolent. Paper presented at the Academy for Eating Disorders International Conference on Eating Disorders, Montreal. Herman, K. M., Hopman, W. M., & Craig, C. L. (2010). Are youth BMI and physical activity associated with better or worse than expected health-related quality of life in adulthood? The Physical Activity Longitudinal Study. Quality of Life Research, 19, 339–349. Hollywood, E., Comiskey, C., Begley, T., Snel, A., O’Sullivan, K., Quirke, M., & Wynne, C. (2012). Measuring and modelling body mass index among a cohort of urban children living with disadvantage. Journal of Advanced Nursing, 69, 851–861. doi:10.1111/j.1365-2648. 2012.06071.x Horowitz, J. L., & Garber, J. (2006). The prevention of depressive symptoms in children and adolescents: A meta-analytic review. Journal of Consulting and Clinical Psychology, 74, 401–415. Houghton, S., O’Connell, M., & O’Flaherty, A. (1998). The use of the children’s depression inventory in an Irish context. The Irish Journal of Psychology, 19, 313–331. Hutchinson, N., & Calland, C. (2011). Body image in the primary school. Oxon: Taylor & Francis. Jansen, W., van de Looij-Jansen, P. M., de Wilde, E. J., & Brug, J. (2008). Feeling fat rather than being fat may be associated with psychological well-being in young Dutch adolescents. Journal of Adolescent Health, 42, 128–136. Judd, C. M., & Kenny, D. A. (1981). Process analysis: Estimating mediation in treatment evaluations. Evaluation Review, 5, 602–619. Kelly, C., Gavin, A., Molcho, M., & Nic Gabhainn, S. (2012). The Irish health behaviour in school-aged children (HBSC) study 2010. Galway: Department of Health and National University of Ireland. Kidscreen Group Europe. (2006). The Kidscreen questionnaires. Quality of life questionnaires for children and adolescents – Handbook. Lengerich: Papst Science Publisher. Knai, C., Lobstein, T., Darmon, N., Rutter, H., & McKee, M. (2012). Socioeconomic patterning of childhood overweight status in Europe. International Journal of Environmental Research and Public Health, 9, 1472–1489. Kovacs, M. (2009). The children’s depression inventory (CDI) technical manual update. Toronto: Multi-Health Systems Inc. Kurth, B. M., & Ellert, U. (2008). Perceived or true obesity: Which causes more suffering in adolescents?: Findings of the german health interview and examination survey for children and adolescents (KiGGS). Deutsches Arzteblatt International, 105, 406–412. Layte, R., & McCrory, C. (2011). Growing up in Ireland: National longitudinal study of children: Overweight and obesity among 9-year-olds. Dublin: Stationery Office.


18

C. Wynne et al.

MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the indirect effect: Distribution of the product and resampling methods. Multivariate Behavioral Research, 39, 99–128. Manios, Y., & Costarelli, V. (2011). Childhood obesity in the WHO European region. In L. A. Moreno, I. Pigeot, & W. Ahrens (Eds.), Epidemiology of obesity in children and adolescents. Prevalence and etiology, (pp. 43–68), Chapter 4. New York: Springer. Meehan, F., Houghton, F., Cowley, H., Houghton, S., & Kelleher, K. (2008). Children’s depression, gender and age norms for an Irish national (primary) school population. Irish Journal of Psychological Medicine, 25, 88–94. Melchior, M., Moffitt, T. E., Milne, B. J., Poulton, R., & Caspi, A. (2007). Why do children from socioeconomically disadvantaged families suffer from poor health when they reach adulthood? A life-course study. American Journal of Epidemiology, 166, 966–974. Musher-Eizenman, D. R., Holub, S. C., Edwards-Leeper, L., Persson, A. V., & Goldstein, S. E. (2003). The narrow range of acceptable body types of preschoolers and their mothers. Journal of Applied Developmental Psychology, 24, 259–272. Muthén, L. K., & Muthén, B. O. (1998–2012). Mplus user’s guide. (7th ed.). Los Angeles, CA. Retrieved from http://www.statmodel.com/ O’Connell, A. (2012). How we see it. Report of a survey on young people’s body image. Dublin: Government Publication, Department of Children and Youth Affairs. Page, R. M., Cole, G. E., & Timmreck, T. C. (1995). Basic epidemiological method & biostat: A practical guidebook. Boston, MA: Jones & Bartlett Learning. Paxton, S. J., Neumark-Sztainer, D., Hannan, P. J., & Eisenberg, M. E. (2006). Body dissatisfaction prospectively predicts depressive mood and low self-esteem in adolescent girls and boys. Journal of Clinical Child and Adolescent Psychology, 35, 539–549. Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36, 717–731. Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42, 185–227. Rajmil, L., Palacio-Vieira, J. A., Herdman, M., López-Aguilà, S., Villalonga-Olives, E., Valderas, J. M., … Alonso, J. (2009). Effect on health-related quality of life of changes in mental health in children and adolescents. Health and Quality of Life Outcomes, 7, 103–103. doi:10.1186/1477-7525-7-103 Ravens-Sieberer, U., Auquier, P., Erhart, M., Gosch, A., Rajmil, L., Bruil, J., … Czemy, L. (2007). The KIDSCREEN-27 quality of life measure for children and adolescents: Psychometric results from a cross-cultural survey in 13 European countries. Quality of Life Research, 16, 1347–1356. Robison, J. (2005). Health at every size: Toward a new paradigm of weight and health. Medscape General Medicine, 7, 13. Rokeach, M. (1973). The nature of human values. New York, NY: Free press. Schaad, A. M. (2012). An examination of the cognitive aspects of the stigma of obesity (Doctoral dissertation). Marietta, OH: Marietta College. Shroff, H, Calogero, R. M., & Thompson, J. K. (2009). Assessment of body image. Handbook of assessment of methods for eating behaviors and weight-related problems: Measures, theory, and research (pp. 115–136), Chapter 4, London: Sage Publications. Siegel, J. M., Yancey, A. K., Aneshensel, C. S., & Schuler, R. (1999). Body image, perceived pubertal timing, and adolescent mental health. Journal of Adolescent Health, 25, 155–165. Stevanovic, D. (2012). Impact of emotional and behavioral symptoms on quality of life in children and adolescents. Quality of Life Research, 22, 333–337. Stice, E., & Shaw, H. (2004). Eating disorder prevention programs: A meta-analytic review. Psychological Bulletin, 130, 206–227.


Psychology & Health

19

Taylor, R. M., Gibson, F., & Franck, L. S. (2008). A concept analysis of health-related quality of life in young people with chronic illness. Journal of Clinical Nursing, 17, 1823–1833. Thompson, J. K., Heinberg, L. J., Altabe, M., & Tantleff-Dunn, S. (1999). Exacting beauty: Theory, assessment, and treatment of body image disturbance. Washington: American Psychological Association. Tsiros, M. D., Olds, T., Buckley, J. D., Grimshaw, P., Brennan, L., Walkley, J., … Coates, A. M. (2009). Health-related quality of life in obese children and adolescents. International Journal of Obesity, 33, 387–400. Turner, E. L., Dobson, J. E., & Pocock, S. J. (2010). Categorisation of continuous risk factors in epidemiological publications: a survey of current practice. Epidemiologic Perspectives & Innovations, 7, 9. Veron-Guidry, S., & Williamson, D. A. (1996). Development of a body image assessment procedure for children and preadolescents. International Journal of Eating Disorders, 20, 287–293. Vitiello, B., Rohde, P., Silva, S., Wells, K., Casat, C., Waslick, B., … Kratochvil, C. (2006). Functioning and quality of life in the treatment for adolescents with depression study (TADS). Journal of the American Academy of Child & Adolescent Psychiatry, 45, 1419–1426. von Rueden, U., Gosch, A., Rajmil, L., Bisegger, C., & Ravens-Sieberer, U. (2006). Socioeconomic determinants of health related quality of life in childhood and adolescence: Results from a European study. Journal of Epidemiology and Community Health, 60, 130–135. Walters, E. E., & Kendler, K. S. (1995). Anorexia nervosa and anorexic-like syndromes in a population-based female twin sample. The American journal of psychiatry, 152, 64–71. Wardle, J., & Cooke, L. (2005). The impact of obesity on psychological well-being. Best Practice & Research Clinical Endocrinology & Metabolism, 19, 421–440. Wesselhoeft, R., Sørensen, M. J., Heiervang, E. R., & Bilenberg, N. (2013). Subthreshold depression in children and adolescents – A systematic review. Journal of Affective Disorders, 151, 7–22. doi:10.1016/j.jad.2013.06.010 Whelton, H., Harrington, J., Crowley, E., Kelleher, V., Cronin, M., & Perry, I. J. (2007). Prevalence of overweight and obesity on the island of Ireland: Results from the North South Survey of Children’s Height, Weight and Body Mass Index, 2002. BMC Public Health, 7, 187. Williams, J. W., Canterford, L., Hesketh, K. D., Hardy, P., Waters, E. B., Patton, G. C., & Wake, M. (2011). Changes in body mass index and health related quality of life from childhood to adolescence. International Journal of Pediatric Obesity, 6, e442–e448. World Health Organisation. (2009). WHO anthroplus for personal computers manual: software for assessing growth of the world’s children and adolescents. Geneva: World Health Organization. Wynne, C., Comiskey, C. M., Hollywood, E., Quirke, M. B., O’Sullivan, K., & McGilloway, S. (2014). The relationship between body mass index and health-related quality of life in urban disadvantaged children. Quality of Life Research, 23, 1895–1905. doi:10.1007/s11136-0140634-7